berlin heart - events-support.com · berlin heart review team . pre-clinical review team . hde...
Post on 19-Jul-2020
4 Views
Preview:
TRANSCRIPT
John Laschinger, MD Medical Officer
CDRH,/ODE/DCD/ICDB
BERLIN HEART
Lead Reviewer
• Shreya Mehta, MS
Medical Officer
• John Laschinger, MD
Mathematical Statistician
• Terri Johnson, PhD
Epidemiology (PAS)
• Veronica Sansing, PhD, MS
Engineering • Brian Fitzgerald
• Michael Long, PhD
• Jean Rinaldi
• Jeffrey Silberberg
• Changfu Wu, PhD
Animal Review • Michael John, MPH
Biocompatibility • Anchal Kaushiva, MS
Sterilization, Shelf Life & Packaging • Lisa Kennell
BERLIN HEART REVIEW TEAM
Pre-Clinical Review Team HDE Review Team
TWO COMPETING PROBLEMS
Lack of effective device therapy for mechanical circulatory support (MCS) in children with acute or chronic heart failure
Lack of monetary incentive for development of a MCS device in orphan population
SOLUTIONS
Berlin Heart Pediatric VAD HDE – Humanitarian Device Exemption
• Alternative Device Regulation Process for Pediatric/Orphan Diseases
• To encourage the discovery and use of devices intended to benefit patients in the treatment or diagnosis of diseases or conditions that affect fewer than 4,000 individuals in the United States.
WHAT IS A BERLIN HEART?
HUMANITARIAN USE DEVICE (HUD) DESIGNATION
Office of Development (OOPD)
Disease manifested in, treats, or diagnoses < 4000 pts./year in US,
total population or
medically plausible subset
No comparable devices available
Device could not otherwise be brought to market
SIGNIFICANT RISK HUMANITARIAN USE DEVICE: HDE PROCESS
Pre-Submission Meeting(s) -Scientific, clinical, regulatory strategies - Review available pre-clinical and clinical data - Pivotal Trial
IDE submission
Pre-Clinical Stage Sufficient safety - Bench testing - Animal Studies - Feasibility/FIM
IDE Approval
MARKET (Site with local IRB)
HDE Submission
HDE Approval
Labeling and SSED
Valid Scientific Evidence of Safety and Probable Benefit
Post Approval Study (PAS)
HUD submission
Pivotal Trial Safety and
Probable Benefit
HUD designation
SAFETY “There is reasonable assurance that a device is safe when it can be determined based on valid scientific evidence that the probable benefits to health from use of the device for its intended uses and conditions of use, when accompanied by adequate directions and warnings against unsafe use, outweigh the probable risks.”
PROBABLE BENEFIT “…the device does not pose an unreasonable risk of illness or injury to patients and that the probable benefit outweighs the risk of injury or illness from its use, taking into account the probable risks and benefits of currently available devices or alternative forms of treatment. .”
SAFETY AND PROBABLE BENEFIT
REGULATORY HISTORY
EXCOR Receives CE Mark
1996 2000
First US EXCOR cases: CU/EU Provisions
2007
EXCOR Receives IDE Approval
2010
FDA Receives EXCOR HDE
EXCOR Receives HUD Designation
2001
UNIQUE PROBLEMS No Approved or Cleared therapy for Comparison Equipoise lacking for randomized comparison to available therapy Data for comparator - ELSO Registry (ECMO) Benefit–Risk Determination complicated by high incidence of stroke (25-30%) deterioration of results with off-label use (stroke
and death)
CONTROLS - VIABLE OPTIONS TO PROSPECTIVE RCT
Prospective, multi-center, single-arm clinical trial Performance Goals Historical Controls Registry based comparisons* Objective Performance Criteria Patient’s as their own control Concurrent Non-Randomized Surgical Controls
* Propensity score matching with concurrent and historical ECMO controls
COMPARATOR FOR EFFECTIVENESS: WHY ECMO (ELSO) HISTORICAL CONTROLS? • No approved or cleared pediatric VADs • ECMO: Current standard of care for pediatric MCS
Not approved or cleared for BTT ELSO Registry – Historical Controls
- Determination of Survival - Propensity matching (n=48 per Cohort, 2:1 matching) - Patients treated after 2000
• Limitations of ELSO registry use: Patients may not be appropriately matched Definition of outcome/recovery/success different for EXCOR and
ECMO ELSO does not capture events after explant (transplantation,
neurologic status, etc.)
• Propensity Score matching – statistical difficulties
DIFFERENCES IN DEFINITIONS
EXCOR ECMO
Endpoint Death on device or survival to transplant or recovery
Death or recovery
Recovery Survival with an acceptable neurological outcome 30 days post explant or to discharge, whichever is later (i.e. successful wean)
Survival for 30 days after being weaned from support
Success Survival to transplant or recovery Recovery
Failure Death, or… Failed wean: •Death within 30 days of explant or prior to hospital discharge, whichever is later •Survival post explant with an unacceptable neurologic outcome
Death on device or within 30 days of being weaned from support
MAJOR INCLUSION/EXCLUSION CRITERIA
Inclusion: • Severe NYHA Functional Class IV (or Ross Functional Class IV for subjects • < 6 years) heart failure refractory to optimal medical therapy • Listed for Cardiac Transplant (UNOS 1A) • INTERMACS profile status 1 or 1A (i.e. critical cardiogenic shock) or • INTERMACS profile status 2 or 2A (i.e. progressive decline) • On ECMO or unable to wean from CPB • Two Ventricle Circulation
Exclusion: • On ECMO > 10 days • Single Ventricle • Hepatic or renal failure (Cr > 3x ULN) • Hemodialysis or Peritoneal Dialysis • Coagulopathy or Platelet disorder
OVERVIEW OF RESULTS – ALL PATIENTS (N= 209)
Total Implants June 21, 2007 – December 20, 2010
n = 204
BSA < 0.7m2 n = 151
Cohort 1 Primary Study
Group n=24
Tx n = 21
Weaned n=1
Death n=2
On Device n=0
Cohort 1 CAP n=20
Tx n = 16
Weaned n=0
Death n=1
On Device n=3
IDE Cohort 3A
CU/EU n=35
Tx n = 20
Weaned n=3 (2/3 Failed)
Death n=10
On Device n=2
Non-IDE Cohort 3A
CU/EU n=72
Tx n = 31
Weaned n=6*
Death n=32
On Device n=3
0.7m2 < BSA < 1.5m2
n = 53
Cohort 2 Primary Study
Group n=24
Tx n = 21
Weaned n=1
Death n=2
On Device n=0
IDE Cohort 3B
CU/EU n=6
Tx n= 4
Weaned n=1
Death n=1
On Device n=0
Non-IDE Cohort 3B
CU/EU n=23
Tx n = 17
Weaned n=0
Death n=3
On Device n=3
PRIMARY EFFECTIVENESS ENDPOINT
16
Effectiveness – Overall Survival: • The objective of the study was to demonstrate that the survival in
subjects treated with the EXCOR Pediatric was different from the survival in the historical control subjects treated with ECMO as a means of mechanical circulatory support (MCS)
Pre-Specified Hypothesis Test:
HR is the expected hazard ratio of EXCOR against ECMO for BTT*
0
1
: 1 vs.
: 1,
H HR
H HR
≥
<
*two-sided α=0.05 using the Cox proportional hazards regression
Primary effectiveness endpoint HR for Cohort 2 cannot be evaluated
– Analysis that appropriately adjusts for matching design cannot be performed since matched controls cannot be identified
– The results for the primary effectiveness endpoint in the HDE may be biased since imbalances in the observed propensity score covariates exist
Unadjusted hazard ratio = 0.043
(P-value=0.004)
Adjusted hazard ratio = 0.099
(P-value = 0.03)
The objective for Cohort 1 seems to be met.
PRIMARY EFFECTIVENESS ENDPOINT – SPONSOR REPORTED COHORT 1 COHORT 2
•p=.0002 vs. ECMO control ** p=.0001 vs. ECMO control
*
**
PRIMARY EFFECTIVENESS ENDPOINT
• Clinical survival and outcome data reviewed: Kaplan-Meier Survival Curves, Competing Outcomes data
• Pre-specified cohorts – separated by BSA • Post hoc cohorts – separated by age (<, or > 4 yrs)
- Did not change composition of primary study groups - New ECMO comparator groups - Provided further information to judge clinical benefit - Data presented by sponsor
• Pre-specified hazard ratio comparison not submitted by the Sponsor
FDA conducted independent analysis FDA Statistical Reviewer conducted FDA analysis on pre-
specified hypothesis test
SURVIVAL ANALYSIS - COHORT 1
COMPETING OUTCOMES COHORT 1
EXCOR ECMO Controls EXCOR ECMO Controls
SURVIVAL ANALYSIS - COHORT 2
EXCOR ECMO Controls
COMPETING OUTCOMES COHORT 2
EXCOR ECMO Controls
Recovered 66.7%
SECONDARY EFFECTIVENESS ENDPOINT CONCLUSIONS
• Days of transplant eligible support (99.3%) • No accounting for real or potential organ
refusal due to temporary clinical status
• Ability to de-intensify concomitant hemodynamic support
• Trend towards decreased levels of support over time was noted
PRIMARY AND SECONDARY EFFECTIVENESS ENDPOINT CONCLUSIONS
• Use of historical controls - problematic • Based upon clinically relevant data
EXCOR provided clinically important benefits for both survival rate and time vs. ECMO
These survival differences are critical for use for a BTT indication • Essential Features of Mechanical Support Device for BTT met:
Survival Prolonged support Preservation of transplant eligibility Ability to de-intensify concomitant support
PRIMARY SAFETY ENDPOINT: ALL IDE PATIENTS Safety:
• The primary objective of the study was to summarize the serious adverse events rate calculated as the number of SAEs per patient-day of support on EXCOR Pediatric.
SAE Rate Performance Goal: • The objective of the primary safety was to compare the SAE rate in
EXCOR treated patients to a performance goal of 0.25 events per patient-day of support
H0: SAE > 0.25 HA: SAE < 0.25
PRIMARY SAFETY ENDPOINT: ALL IDE PATIENTS
* < 0.25 SAEs per patient day of support
*
*
STATISTICAL SUMMARY
• Primary safety objective was met • Primary effectiveness objective
Cohort 1: seems to be met; however, results may still be biased due to imbalances in omitted covariates from the propensity score model (BSA)
Cohort 2: statistically inconclusive; in addition to imbalance in omitted covariates, imbalances in observed covariates still exist
Clinical judgment is needed
FDA REQUESTED ADDITIONAL ANALYSES
FDA requested additional data from the Sponsor in four key areas to further understand risk/benefit profile
I. Mortality in Non-Primary Implant Groups II. Stroke and Neurologic Status and Outcomes III. Health Related Quality of Life Data IV. Pump Thrombus – not an AE but common
ADDITIONAL FDA CONCLUSIONS MORTALITY • Low mortality is observed when:
• Device is implanted at experienced centers and • Requirements for use include strict inclusion/exclusion criteria
• Pre-Implant Predictors of mortality included: • Single ventricle circulation
• Any pre-implant use of ECMO
• For all CU/EU patients • Total mortality and failed wean:
- Was high
- Was not affected by the site of implantation (IDE vs. non-IDE)
SURVIVAL AND NEUROLOGIC OUTCOME: COHORT 1
Cohort 1 (n=24) 21Transplanted (87.5%)
2 deaths 1 Failed Wean
Alive and Transplanted with Good* Neurologic
Outcome (70.8%)
Normal No/Mild n=15 n=2
Alive and Transplanted with Poor** Neurologic
Outcome (16.7%)
n=4
Death or Failed Wean (12.5%)
Neuro Other Injury Cause
n=2 n=1
* Good = PSOM <1; **Poor = PSOM > 1
SURVIVAL AND NEUROLOGIC OUTCOME: COHORT 2
Cohort 2 (n=24) 22 Transplanted or Weaned (91.7%)
2 deaths
Alive and Transplanted or Weaned with Good* Neurologic Outcome
(75%) Normal No/Mild
n=16 n=2
Alive and Transplanted with Poor** Neurologic
Outcome (16.7%)
n=4
Death (8.3%)
Neuro Other Injury Cause
n=2 n=0
* Good = PSOM <1; **Poor = PSOM > 1
SUMMARY OF NEUROLOGIC OUTCOMES IN STUDY COHORTS • Survived to transplant or successfully weaned with no
neurologic events 64.5% (15/24) of Cohort 1 66.7% (16/24) of Cohort 2
• Survived to transplant or were successfully weaned with either no neurologic events or a good neurologic outcome (PSOM <1). 70.8% of Cohort 1 (17/24) 75% of Cohort 2 (18/24)
35/48 = 72.9%
SUMMARY OF CLINICAL REVIEW • Primary safety was met based upon pre-specified hypothesis • EXCOR showed probable benefit based on submitted data with clinically important improvements in the ability to provide prolonged support
• Pump performance: No device failures Excellent long-term hemodynamic support allows use as BTT Concomitant support able to be weaned in selected patients Prone to thrombus formation – inlet and outlet valves
• Results indicate increased mortality can be expected in certain situations: Broader clinical use for BTT indication Relatively less-experienced centers Patients with single ventricle physiology Patients requiring pre-implant ECMO support
AREAS REQUIRING ADDITIONAL STUDY - PAS • Neurologic Events and Outcome
Risk of ischemic neurologic events is uniformly high (25-30%) Almost all deaths/ failed weans or poor outcomes – Neuro Events Long term effects on neurocognitive function unknown • Overwhelming majority (72.9%) - transplant or successful weaning with either no neurologic events or good neurologic outcome
• Health related quality of life Worse than levels typically observed in chronic disease states Only measured on device
• Pump thrombus is a significant problem Likely related to pump materials and/or design Acute effects include:
- Higher incidence of ischemic stroke - “Poorer” neurologic outcome - Lower HRQoL assessments
top related